Reading instruction prompts the emergence of neural circuits that are specialized for rapidly translating printed symbols into sound and meaning. Understanding how these circuits differ in children with dyslexia, and change with learning, is an important scientific challenge that holds practical implications for education. The proposed research employs longitudinal measurements in children with dyslexia over the course of a highly effective intervention program to: (a) determine how brain structure and function change in response to reading instruction; and (b) investigate neurobiological factors that predispose a child to struggle or succeed in the intervention. Thus, this proposal seeks to determine both how education shapes brain development, and how a child?s unique neurobiology predicts educational outcomes. Aim 1) White matter plasticity and learning ? White matter was previously considered static infrastructure; it is now known that many cellular properties of the white matter change with learning. Reading interventions provide a powerful tool to study experience-dependent plasticity in the human brain. Based on novel quantitative MRI and diffusion MRI measurements developed by the PI and collaborators, it is now possible to quantify changes in cell density, intra-axonal water and myelination at millimeter resolution. Longitudinal measurements collected before, during, and after the intervention will be used to model the time-course of white matter plasticity associated with improvements in reading skills, and investigate the biological mechanisms that underlie differences in learning among children. Aim 2) Bottom-up and top-down computations in the reading circuit - When our eyes fixate upon a word, a cascade of neural processes is initiated, beginning in the visual system and progressing through a series of computations that translate the visual representation into sound and meaning. We have developed the first model of the neural computations performed by critical components of the brain?s reading circuitry. Based on this new understanding of computations in the literate adult brain, we will investigate how computations differ in children with dyslexia, and change in response to reading instruction. Aim 3) Neural biomarkers of learning outcomes - Even in a controlled and intensive learning environment, some children show substantial improvements in reading skill, while others show limited change. What biological factors predispose a child to excel or struggle when provided a high-quality intervention? Pre- intervention MRI measurements will be examined as predictor variables for individual differences in intervention learning rate, and long-term, post-intervention outcomes. Aim 3 will capitalize machine learning to develop a model of the neuroanatomical factors that predict learning outcomes.